GP-97 Conference Report

The Second International Conference on Genetic Programming (GP-97), like the first, was held on the beautiful Stanford University campus in California under the chairmanship of John Koza. More than 350 people from all over the world gathered together for four days to see presentations, posters, tutorials and trade presentations on a range of topics. In addition to GP and related topics in Genetic Algorithms (GAs) and classifier systems, the up and coming fields of evolvable hardware (EHW) and biocomputing (also called DNA computing) were also represented.

Each day kicked off with a keynote presentation by invited speakers renowned in their field, such as David Fogel. These were followed by plenary sessions including papers in each of the conference topics. The afternoons where given to parallel paper presentation sessions specialising on particular topics.

Of the 20 tutorials on offer, perhaps the stand out was Russ Altman's Molecular Biology for Computer Scientists. This highly interactive whistle stop tour of DNA, RNA, amino acids, proteins, genes etc. earned him the title of "Jedi Master of handling hecklers" and left his audience enthused with the biology from which GP draws its inspiration.

Registration included use of Stanford campus facilities. Of these the library (with its internet PCs), book shop (giving a 10% discount to conference registrants), gyms, swimming pools, art gallery and Rodin garden proved popular, while the bars and restaurants of the student union and adjacent downtown Palo Alto served as locations for animated discussions and exchange of ideas (many of them related to GP !)

The two ends of the PhD process where both represented at the conference. The day before GP-97 started a workshop for current PhD was held and the final plenary session finished with a panel discussion by holders of PhDs in GP.

The international aspect of GP-97 was reflected by the decisions to hold the inaugural meeting of the GP research group of the European union, EvoGP, at the conference and the first editors' meeting for the follow up to Advances in Genetic Programming 1 and 2 during the conference.

While technical papers included those concentrated on schema, fitness, extracting data from huge volumes of data or medical measurements, perhaps the most entertaining presentation showed robots learning to move by swinging hand-over-hand like a gibbon (Hasegawa and Fukuda).

The third annual GP conference (GP-98) will be held at the University of Wisconsin in Madison, USA 22-25 July 1998 (just before AAAI-98 which will also be in Madison) and the first European workshop on GP will be held in Paris in the Spring. They will find GP-97 a tough act to follow!


W. B. Langdon, The University of Birmingham